Effective cloud orchestration involves automated software-based discovery, monitoring, provisioning and optimization of complex IT infrastructure and IT service management processes. Software-defined-X is sweeping the technology landscape by separating the data-plane from the control-plane. This provides greater flexibility to the dynamic usage patterns of cloud resources. We’ve built industrial-grade cloud orchestrators encompassing network management, e-discovery and ITSM (IT Services Management) along with control-plane solutions to accelerate cloud management and orchestration requirements.
Working with Zymr surpassed our expectations. Zymr leveraged deep competencies to build a highly scalable, multi-tiered architecture for application performance management, fault management and correlation, for our cutting-edge Datacenter Operations Analytics Suite. I recommend them highly to anyone seeking reliable and agile product engineering.Sameer Padhye CEO – FixStream Networks
Every new software service, for example SaaS products, need a comprehensive cloud orchestration and management solution. This starts with a well defined ontology and object models of the service. Next is designing the life-cycle from initiation to tear-down. Initiation requires well designed automated service provisioning, authorization and billing. During normal operations it is vital to monitor for performance and faults in order to optimize the service for SLA compliance.
Cloud, enterprise and carrier service provisioning all require well thought out ontology and object models. Provisioning is done through well structured APIs which are optimized for automation. Service management requires seamless integration with subscriber billing and accounting.
Proper instrumentation and logging of the service is the prerequisite for service monitoring. Only monitoring instances of cloud VMs like Amazon AWS’ CloudWatch is a good start, but should be complemented by instrumenting the innards of the service delivery engine for deeper resolution.
Strong DevOps processes enable the efficient and automated upgrading of software hosted on public or private clouds. It is important to integrate automated DevOps from the very start using methods like Puppet, Chef, Ansible or Python scripting. We encourage the containerization of services using new paradigms like Docker.
Networks are no longer islands. Managing networks effectively requires a comprehensive understanding of the underlying networking infrastructure including FCAPS. Classic network management layers comprise of EMS, NMS, OSS and BMS. With the advent of SDN, network management has morphed to cloud-based software controllers that need an accurate view of the underlying network topology to provision flows, and monitor and adapt based on QoS to recover from faults.
Discovering networking assets to gauge their connectivity, configuration and operational status is essential to getting a handle on the network topology. We leverage our experience in building discovery technology accelerators as well as products for clients in the EMS, NMS and Cloud Controllers space.
Once network topology has been established and networking assets have been discovered. The next step is implementing FCAPS with advanced features like rules based fault correlation and policy management.
NMS has evolved from passively monitoring dashboards to fully automated policy-based orchestration solutions. Automatic responses prevent network failures by integrating active monitoring and response to faults like SNMP Traps. We develop automated cloud orchestrators using streaming analysis of faults and triggering BPM-based workflows to autocorrect concerns before they become crises.
The volume of data and its attendant storage is exploding. The challenge only grows with storage innovations like virtualization, Tier-0 SSD for faster IOPS and data durability, the transformation of object storage and technologies such as Amazon’s S3 and OpenStack Swift. Solutions must incorporate new standards from SNIA for CSMI, including SNIA’s SMI-S or SMI-C management profiles to manage storage based on APIs.
SNIA’s CSMI may evolve into a unifying format to manage the delivery of virtualized storage on demand. Leveraging CSMI’s RESTful APIs can enable cloud storage management as an overlay, or integrated as part of a newer cloud storage service.
SNIA’s standardization of management through SMI-S allows vendors to enable the management of their storage devices using reference models. We develop SMI-S interfaces to support various standards and proprietary storage management profiles.
Exposing management interfaces using SMI-S or similar APIs is good, but some enterprise and service providers may require storage management to be integrated with their DCIM and other NMS or OSS systems.
DCIM has always been very complex due to frequent upgrade cycles and multi-vendor infrastructure. Today’s cloud based services–IaaS, PaaS, SaaS–are provided by virtualized services. There is also an increasing emphasis on improving datacenter efficiency through metrics tracking power consumption, bare-metal server workloads and VM, network, and storage management. We develop DCIM solutions from scratch to cover energy, server, network, storage, hypervisor, OS and application monitoring.
DCIM solutions are generally built with an accurate CMDB that includes auto-discovery of heterogeneous datacenter assets. Monitoring and optimizing these resources requires keeping track of and updating datacenter inventory. We build comprehensive multi-vendor auto-discovery services for DCIM and record these assets in a CMDB based on ITILv3 standards.
Developing metrics for operational efficiency starts with monitoring datacenter assets, virtual services and applications. Different asset classes often have their own protocols. Networks use SNMP, storage uses SMI-S and hypervisors have their own set of APIs. Application monitoring is also multi-faceted, with some Java applications providing JMX APIs with others using proprietary APIs. We build comprehensive multi-vendor and multi-protocol monitoring systems within our cloud orchestrators.
Big data analytics makes it possible to integrate real-time streams of threshold crossing events. Analysis of these streams can lead to further optimization through automated orchestration. For example, recovery procedures can be automatically kicked in when the availability zone of a datacenter is performing poorly. We build automated cloud orchestration solutions based on user-defined workflows to integrated fault and performance monitoring events.
Work with the same high quality software development teams that have worked with many successful Silicon Valley startups, and enterprises. We have been a partner of choice for some of the most discriminating tech leaders and tier-1 venture capitalists due to our cutting-edge skills, agility and attitude.
Let Zymr help you enhance your core solutions to accelerate your cloud roadmap.
Smartsourcing is a brief guide to the world of modern technology partnerships. It was developed through a collaborative effort of top Zymr executives as we uncovered a gap in the market between the perception of what outsourcing used to be, and how leading technology innovators are leveraging this globalized approach to value generation. Read this guide to learn...